境外移入病例與社會人口因子在登革熱疫情空間擴散的角色

Abstract

登革熱為近三十年在全球擴張的蚊媒性疾病，伴隨全球暖化所帶來的擔憂，是此類熱帶性疾病於亞熱帶甚至溫帶地區的流行病區在地化成常年病，因此極需瞭解流行病區境外移入與本土流行的關係。本研究以臺灣登革熱流行病區做為先驅研究，在日益蓬勃的航空運輸中，抬升旅客從東南亞等常年病區攜入登革熱病毒的風險，誘發臺灣流行的爆發。既有研究已指出境外移入案例為登革熱流行爆發的點火器，但過往資料在解釋登革熱傳遞特性側重於氣象資料，而忽略人類活動的介入，使得氣象資料的解釋不夠完備，因此對於誘發本土案例出現的時空關係仍然不清楚。本研究旨在使用台灣各鄉鎮的社經條件和特性，對境外移入案例誘發本土案例的傳遞風險和時間延遲長度進行分析，以細緻化境外移入案例在空間上和時間上的異質性影響效果。 本研究透過2003至2012年臺灣各鄉鎮每週境外移入案例數與本土案例數，建構一基礎分析單位，傳遞的起終點配對(Origin-Destination pair)，並依傳遞關係區分出境外移入鄉鎮及本土案例出現鄉鎮兩類別，每個類別皆選用臺灣鄉鎮發展類型、社會經濟指標、是否位於潛在傳染區為解釋變數，再進行多元迴歸分析，依全域模型和分月模型對境外移入對向外傳遞風險與傳遞速度進行討論。 研究發現，在境外移入傳遞風險在流行季節(7至12月)經距離倒數的權重調整後，社經條件即具備解釋能力，都市化程度高但收入低之境外移入鄉鎮對外傳遞的風險較高。在傳遞速度上，社經條件只在晚春（3至4月）具備較高解釋力，高人口密度但收入低之境外移入鄉鎮和高人口密度且高收入本土案例鄉鎮的組合加速傳遞速度，然而在初夏（5至6月）境外移入鄉鎮和本土案例鄉鎮的高人口密度反而延緩了傳遞速度。透過時間延遲預測地圖顯示，晚春（3到4月）需要預警的地方主要集中在高度都市化與中度都市化之鄉鎮。初夏（5到6月）時，都市化程度高之鄉鎮的其本土案例的時間延遲聚集效應消失。 本研究指出境外移入案例出現的所在地，其在地的人口特性、經濟條件和都市類型能決定本土案例出現的風險，且這些影響效果跼限於特定的季節，待未來研究進一步指出內部原因。本研究結果可提供疫政單位，在境外移入案例出現時，提前預警可能出現本土案例之鄉鎮及時間，而有效的進行防治工作，阻斷登革熱疫情的爆發。

Dengue fever, a mosquito-borne disease, is an expanding public-health problem worldwide. One major research question involves the geographic expansion of such tropical diseases with global warming. Therefore, it is important to examine the relationship between imported cases and indigenous cases in epidemic regions. This pilot study took place in Taiwan. Air travel has brought more travelers from a dengue-endemic region, Southeast Asia, who have introduced virus strains that increase the risk of local epidemics in Taiwan. Past research has concentrated on meteorological data and ignored human behavior and socioeconomic conditions that influence dengue transmission at the meso- or micro-scale. The objective of this study is to examine the influences of socioeconomic conditions on the risk and speed of transmission. This study used origin-destination (OD) pairs to measure the risk and speed of transmission as dependent variables. Each OD pair was taken from cases of dengue fever in Taiwan between 2003 and 2012. The multiple-regression analysis included socioeconomic indicators, urbanization type, and past record of local emergence, as explanatory variables. The results showed that the risk of diffusion is correlated with the socioeconomic status of imported townships and with inverse distance between locations between July and December. Low average income in urbanized townships raised the risk of transmission of imported townships. Transmission speed was best explained by socioeconomic status of imported and local townships in late spring (from March to April). Higher population density in both imported and local townships was associated with increased transmission speeds, as was the pairing of low-average-income imported townships with higher-average-income local townships. In early summer (from May to June), higher population density was associated with lower transmission speeds. A predicted distribution of time lags shows the importance of urbanization with short time lags in late spring, though this effect disappeared in early summer. This study indicates that socioeconomic status of townships would influence transmission; however, the effect is limited to particular seasons, and the drivers are still unclear. Future research should focus on clarifying these relationships.